跳到正文
arXiv cs.AI· Mingyang Liu, Qingcan Kang, Shixiong Kai, Kaichao Liang, Xiaojin Fu, Tao Zhong, Linqi Song·· 4 小时前AI 评分38

EvoDisco:面向演化式智能体记忆的策略发现

EvoDisco: Strategy Discovery for Evolving Agent Memory

AI 导读

EvoDisco 将智能体记忆演化建模为演化式策略发现,由可学习策略根据当前记忆状态与近期任务历史,从 Abstract、Refine、Bridge、Reframe 四种语义算子中选择并实例化,而非按固定算子调度重写记忆。

正文 · 原文

View PDF HTML (experimental)

Abstract:Large language model agents accumulate experience in memory, and how that memory evolves during interaction is as important as what it stores. Recent methods update memory online, yet both the operators that transform memory and the rules that select them remain fixed by design. We propose EvoDisco, which formulates agent memory evolution as evolutionary strategy discovery: the agent must discover which operator to apply, when to apply it, and how to instantiate it, rather than rewriting memory under a static operator schedule. A learnable policy conditions on the current memory state and recent task history, then selects and instantiates one of four semantic operators---Abstract, Refine, Bridge, and Reframe---which respectively compress related memories, decompose overloaded entries, connect weakly linked experiences, and reorganize retrieved reasoning chains. The policy is updated with a bi-level fitness. An inner signal scores each transformation using a retrieval-quality improvement regularized by memory complexity. An outer signal scores strategic efficacy over a window of past decisions, combining historical operator-utility regret, utility-prediction error, and selection diversity. Under a one-pass online plus frozen-transfer protocol on ALFWorld, WebShop, and LifelongAgentBench (OS and DB), EvoDisco improves over strong memory baselines including MemP and MemRL. The largest gains are on ALFWorld ($+8.8$ online / $+12.9$ frozen-transfer points over MemP; $+16.4$ / $+17.9$ over MemRL). Ablations attribute these gains to both fitness levels and the operator library.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.22102 [cs.AI]
  (or arXiv:2611.22102v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.22102

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Mingyang Liu [view email]
[v1] Tue, 17 Nov 2026 02:37:56 UTC (1,377 KB)

来源:arXiv cs.AI · arxiv.org